Transcript High frequency power spectrum of Cygnus X-1 from the USA

SLAC’s First Step into Space:
Status of the USA experiment
5-years after launch
Larry Wai
SLAC / Group K
The beginning of particle
astrophysics at SLAC
Experimental struggles and the
rewards of science
Outline of talk
1.
The life and times of the USA experiment
• The people who made it work
• On-orbit adventures
• The detector and its calibration
2. Science from the USA experiment
• Tests of general relativity in black hole systems


•
Physics of “jets” in black hole systems

3.
The high frequency power spectrum of Cygnus X-1
A search for x-ray bursts from ~10 solar mass compact
objects
Flares In BL LAC object 1ES1959+65
Summary and plans
Larry Wai, SLAC seminar
3
Part 1: The life and times of the
USA detector
• 1991-1998. Design,
manufacturing,
integration, testing,
calibration, storage
(satellite late)
• T0 = February 23, 1999;
Delta-II launch from
Vandenberg AFB, CA
• End of USA mission at
T0+21months (3 months
shy of design lifetime of 24
months)
Larry Wai, SLAC seminar
4
The people who made it work
How SLAC students and staff got
their hands into a space based
experiment and made it fly
USA Collaboration
USA
USAX-ray
Telescope
(1-16keV)
NRL: R. Bandyopadhyay, G. Fritz, P.
Hertz, M. Kowalski, M. Lovellette (P.S.),
P. Ray, L. Titarchuk (& GMU), M. Wolff,
K. Wood (P.I.), D. Yentis, W.N. Johnson
SLAC/Stanford: E. Bloom (S.U. Lead
Co-I), W. Focke, B. Giebels, G. Godfrey,
P. Michelson, K. Reilly, M. Roberts, P.
Saz Parkinson, J. Scargle ( & NASA
Ames), G. Shabad, D. Tournear
Larry Wai, SLAC seminar
6
The USA project - pushing the limits
SLAC contributions:
~1/5 the
manpower
of Shabad (Ph.D.
•Detector
calibration
Gary
Godfrey,
Ganya
• SLAC mechanical
and~½
thermal
design,
validation GLAST
in
the
time!
Physics),
Pablo
Saz-Parkinson
(Ph.D.
Physics)
,
Berrie
John Hanson (Ph.D. Aero-Astro), Alex Leubke
Giebels
(M.S./Engineering Aero-Astro)
•Ground software - Kaice Reilly (Ph.D. App.Physics), Derek
Tournear
Physics), Warren
Focke framework,
• SLAC(Ph.D.
manufacturing
of mechanical
•Science
operations
- Student
& post-doc
involvement on a
collimators
– John
Hanson,
John Broeder
weekly basis deciding what sources should be observed
• Flight
software
– SLAC
contributions
(many
students
defined
the subject
of their Ph.D. science in
this way); heavy involvement in all the publications
Larry Wai, SLAC seminar
7
On-orbit adventures
How USA did it the hard way and
made it work
The pointing challenge
Thechallenge:
USA innovation:
The
mechanical
rotation
to point detector
at celestial
by
••Use
Spacecraft
(ARGOS)
axissystem
was continuously
oriented
normal sources
to earth’s
setting
yaw
(X-axis rotation),
in pitch
(Y-axis
surface
throughout
the orbitfollowing
– need tosource
keep USA
pointed
onrotation)
a celestial
point source to ~0.05 degrees
Celestial source
Yaw source
Celestial
ARGOS
USA
Celestial source
USA
USA
ARGOS
Pitch
ARGOS
Earth’s center
Earth’s center
Earth’s center
Larry Wai, SLAC seminar
9
10,000 orbits
The polar orbit challenge
USA reached 87% of design lifetime on-orbit
•Unusual USA characteristic - equatorial orbits
During
lifetime
detector on-orbit
are
preferable
forof
astronomy
(backgrounds are
•10% time used for satellite calibration
better)
•14% time
used solving
pointing by
problems
•Orbit
was divided
into segments
passage
(due to earth’s
satelliteradiation
misinformation
through
belts andon
the South
orientation;
USA
diagnosed
the
problem)
Atlantic Anomaly
•76%20min
time had
good pointing
•Two
(equatorial)
and two 10min (polar)
observations segments per orbit
Larry Wai, SLAC seminar
10
Celestial source observations
Source Name
Crab_Pulsar
X1630-472
Cyg_X-1
Cyg_X-2
XTE_J1118+480
SMC_X-1
E_2259+586_SNR
Cas_A
Circinus X-1
Mkn_421
Cen_X-3
X0614+091
GRS_1915+105
GX_349+2
X0142+614
ksec
1220.
716.6
706.0
626.6
601.1
415.4
397.4
370.0
345.3
278.2
265.7
249.0
243.6
232.9
224.9
comments
pulsar
black hole
black hole
neutron star
black hole
pulsar
pulsar
supernova remnant
neutron star
active galactic nucleus
pulsar
neutron star
black hole
neutron star
pulsar
•Strategy: accumulate long object
observing times ~fraction of a
month
•About 90 Sources Observed by
USA
•Top sources in observing time
had 0.1-0.5 months each!
Larry Wai, SLAC seminar
11
The detector and its calibration
A story of careful work on the
ground, an on-orbit surprise, and a
lesson learned
Detector design goals
• Low energy threshold (~1 keV)
• Large collecting area (~2000 sq.cm)
• High time resolution (~1 microsecond)
• Sustained high data rates (40-128kbps)
Larry Wai, SLAC seminar
13
X-ray detection technique: collimator +
standard multiwire proportional chamber

X-ray photon
Collimator ~ 1.3o FWHM
~10% of 1KeV
photons pass
through
pressure
Gas Volume (P10)
window
90%Argon
10%Methane
50 m wire
anode
Initial Ionization
40-600 e-ion pairs
2 sec time resolution
(typical ~ 32 sec res)
E/E ~ 17% at 6 keV
~105 electrons/event
Larry Wai, SLAC seminar
14
USA detector details
USA X-ray Detector (1 of 2 Identical Modules)
USA Proportional Chamber
Kapton(2.5 um) + Aluminum(.1 um)
Sun Shield
Top View
2.8 cm
29"
End View
Support Mesh (85% Transmission)
Mylar (5.0 um) + Nichrome (.01 um)
Pressure Window
1.3 Deg
FWHM
12.5"
4.5"
Strongback
Collimator
29" wires (2)
Periphery anode
for charged veto
Copper
Proportional
Chamber
2.8 cm
Thin Window
2 interleaved wires running serpentine through each
Diam
layer12.5"
x 2 layers + 1 wire around the outside for ~1/8"
(point to point)
anticoincidence
Larry Wai, SLAC seminar
15
Effective area determination
Transmission of
m Mylar
+ 2.5 m Kapton
• ~200’ long tube w/ 55Fe source
at5.0
one
end,
Absorption edge of Argon
collimator
on rotation
•Energy
dependence
of fixture at other; ~1Hz count
rate through
collimator
effective
area derived
from
Livermore
cross- vs angle of incidence (point
• Measurex-ray
acceptance
section
formulae
forcollimator
various effective area ~1000 sq.
spread
function,
detector
cm) materials
•Superior effective area
below 4 keV as compared to
PCA (proportional counter
array aboard RXTE – NASA
mission up since Dec. 1995)
Larry Wai, SLAC seminar
16
T-Vac high rate tests
16.5 s
•
55Fe
source fastened to
yoke, scan in yaw
• Ganya fitted histograms of
time difference between
events with event time
domain model including
deadtime and other
electronics effects
• Extraction of deadtime as a
function of rate
55Fe
source mounted to yoke
Larry Wai, SLAC seminar
Detector 1
1035 cts/sec
c2 / dof = 0.979
DOF = 493
17
Power spectrum tests
General procedure for power spectrum:
Leahy
normalized
power
spectrum
••Power
Break down
all data into equal
length the
time segments
(T),
spectrum:
convert
time
series
USA calibration data:
each with N equal length bins
deadtime introduces
of
counts
into
the
frequency
domain
Purely
Poisson
process
• For
each segment
calculate the “Leahy normalized” power
correlations
between
2 GoodPagreement
2between
spectrum
=2|X
|
Xj is the
amplitude of the
j
j /Ncounts where Good
photon
times
agreement
deadtime model and data
discreteusing
Fourier
transform:
Rate
= 4075
all energy
calibration
data
•Basic
the test: between
check
forcts/sec
subtle
power idea forchannels!
Pj=PGanya’s
and
deadtime
1+P2cos(2pj/N)
systematic effects in themodel
data
P1calibration
=1.763,
P2=-0.0245
– when
using
all
c2 / dofchannels
= 1.08
energy
DOF = 2046
combined
th time bin
xk (k=0,1,…,N-1)
is
the
number
of
counts
in
the
k
0
0 segmentsFrequency
• Average
to get mean and RMS
5200 Hz
Larry Wai, SLAC seminar
18
An on-orbit surprise
• Pablo notices recurring
patterns of distortions in
the power spectrum of
celestial sources when
selecting energy bins
Energy
Energy channel
channel 12
Larry Wai, SLAC seminar
19
Going back to the calibration data
• Ganya goes back to calibration data and confirms
an energy dependent instrumental effect (a.k.a.
EDIE) on power spectrum
• Positive vs negative spectral slope (as measured in
detector) inverts shape in frequency domain
• Effect cancels out when all energy channels are
combined; that’s why it was missed during T-Vac
testing
• Working hypothesis is pulse tail oscillations;
phenomenological corrections used at present
Larry Wai, SLAC seminar
20
A lesson for space-based detectors
• More manpower and time in analysis of data
during detector testing on the ground could
have unearthed EDIE before launch and
allowed us to characterize the effect more
carefully than was possible on-orbit
• The lesson learned: let’s check the data
carefully during testing of the GLAST large
area telescope at SLAC! (2004-2005!)
Larry Wai, SLAC seminar
21
Part 2: Science from the USA
experiment
Selected Astrophysical Journal papers:
• USA and RXTE Observations of a Variable Low-Frequency QPO in XTE
J1118+480, K. S. Wood et.al. , ApJ (2000)
• Disk Diffusion Propagation Model for the Outburst of XTE J1118+480,
Kent S. Wood et al., ApJ (2001)
• USA Observation of Spectral and Timing Evolution During the 2000
Outburst of XTE J1550-564, K. T. Reilly et.al., ApJ (2001)
• Eclipse Timing of the Low Mass X-ray Binary EXO0748-676 III. An
apparent Orbital Period Glitch Observed with USA and RXTE, M. T.
Wolff et.al., ApJ (2002)
• Observation of X-ray variability in the BL Lac object 1ES1959+65, Berrie
Giebels et.al., ApJ (2002)
• X-ray Bursts in Neutron Star and Black Hole Binaries from USA and
RXTE, D. Tournear et.al., ApJ (2003)
• High frequency power spectrum of Cygnus X-1 from the USA
experiment, W. Focke and L. Wai et.al., (in progress)
Larry Wai, SLAC seminar
22
High frequency power spectrum
of Cygnus X-1
Testing a prediction of General
Relativity: the innermost stable
circular orbit
An innermost stable orbit
Relativistic
Effective
Potential
Stable circular orbit
L
 1.875
mcRSch
Distance from center of black hole
• Innermost stable circular orbit (ISCO) at slightly
more than a x (M/Msolar)3km, 2<a<4.5
Larry Wai, SLAC seminar
24
Looking for the innermost stable
orbit in Cygnus X-1
• Cygnus X-1: aorbiting
~10
•Non-uniform
solarinmass
hole
matter
the black
disk will
candidate with a
produce
variations
in
companion
star
•Signature of the
observed
fluxmatter
at orbital
donating
to an
innermost
accretion stable
disk around
frequency
the black
circular
orbit
is aand
sharp
•Models,
e.g.hole
Bao
• X-ray luminosity
from
drop-off
in thepredict
power
Ostgaard
(1995),
Cygnus X-1 originates
-1 up
power
spectrum
P~f
spectrum
at
~220Hz
in ~10KeV plasma
to the
frequency“seed
of the
upscattering
photons”stable
fromorbit
innermost
orbiting
matter in the
fISCO
= (Msolar/M)2.2kHz
disk
Mass-Donor Companion Star
~106 km
~103 km
Accretion
Disk
Larry Wai, SLAC seminar
Black hole
25
Extracting the power spectrum
• For each time segment (~1sec) calculate the power
spectrum and subtract the noise including deadtime
distortion (from Ganya’s results)
• Average all the resulting power spectra over all
segments (~400k)
• Fit in region above 2kHz to correct for residual
noise/deadtime
• Fit in region above 300Hz to correct for residual
EDIE
Larry Wai, SLAC seminar
26
Cygnus X-1 Power Spectrum
•Model the “drop-off” as a broken power law
•Best fit broken power law has c2=1457 for 1437 DOF
•Best fit single power law has c2=1465 for 1439 DOF
c2=8 with 2 additional degrees of freedom
•2.5 sigma effect – marginal evidence for a dropoff
f-1.6
Focke, Wai,
Bloom, et.al.
Residual EDIE
Residual deadtime
Larry Wai, SLAC seminar
27
A search for x-ray bursts from
10 solar mass compact objects
Testing another prediction of
General Relativity: the event
horizon
The measured masses of compact
objects
•Maximum
• In a binaryneutron
system,
starneed
massorbital
is 3.2period,
solar
velocity, partner mass,
masses
and angle of
•Sample
of observed
inclination
to estimate
~10
thesolar
massmass
of theobjects
object to
are compact
widely believed
• black
Two populations
be
holes - with an
emerge, one around
event
horizon at
~1.4 solar masses, and
(M/Msolar)3km
~10 solar masses
Neutron star mass limit
Neutron Stars
Black Hole Candidates
Sco X-1
Cyg X-2
Larry Wai, SLAC seminar
Cyg X-3
XTE J1550-564
XTE J1859+226
GRS 1915+105
XTE J1118+480
Miller (1998)
+Tournear (2003)
29
Using bursts as an event horizon
litmus test
• Observation of thermonuclear burning on the surface of
the black hole candidate would reject the event horizon
hypothesis
• The signature: type 1 x-ray bursts
• These bursts are due to unstable thermonuclear burning on
the surface of neutron stars (cooling blackbody
temperature, radiating area corresponding to 10-15km
radius sphere, and linear correlation between burst flux and
time delay)
• Narayan-Heyl (2002) prediction for bursting luminosity
region for 1.5 and 10 Msolar compact object w/ baryonic
surface
Larry Wai, SLAC seminar
30
Black hole candidate burst rate limit
Tournear,
Bloom, et. al.
(2003)
• Result: BHC burst rate is less than 5% of that for
neutron stars (at 95% C.L.)
• Black hole candidates quantitatively don’t have
baryonic surfaces!
Larry Wai, SLAC seminar
31
Flares In BL LAC object
1ES1959+65
Testing a prediction about how an
AGN jet works
Black holes, small and large
Active galactic nuclei
(AGN)
• ~106-9 solar mass
black hole
• ~109 km disk
• Jets of electrons!
• E.g. 1ES1959+65
•GLAST bread
and butter
Large!
Small!
Galactic black hole
Microquasar
candidate
• ~10 solar mass
black hole
• ~103 km disk
••E.g.
Jets!
Cygnus X-1
•E.g. GRS 1915+105
~106-9 solar
mass black
hole
~10 solar
mass black
hole
Larry Wai, SLAC seminar
33
USA AGN observations
We analyzed
this one so
far…
Larry Wai, SLAC seminar
34
Flaring in BL Lac object
1ES1959+65
• 2000 Sept-Nov.
observation of
variability by USA
led to search in TeV
• 2002 May-July
observations by
Whipple of clear
TeV gamma ray
flaring
Giebels, Bloom,
et.al. et.al.
(2002)
Holder,
Daily x-ray flux
(2003)
daily TeV
gamma flux
Hardness ratio
Larry Wai, SLAC seminar
35
Confirmation of a prediction
• 1ES1959+65 was predicted to be the 3rd brightest
extra-galactic TeV source by Stecker et.al. (1996)
based upon data from the two known extragalactic
TeV sources Mrk 421 and 501
• Prediction based upon “synchrotron self compton”
scattering in AGN jets as the mechanism for TeV
emission
• x-rays come from synchrotron radiation of jet
electrons, and TeV gammas are the x-rays
Compton upscattered by the same jet electrons
• Example of a multiwavelength campaign (which we
will need in GLAST to study jet physics)
Larry Wai, SLAC seminar
36
Summary and plans
How did we do, and what is left?
How did we do?
SLAC’s first step into space
SLAC people contributed to the design,
made flight hardware/software, tested and
calibrated the detector, helped define the
observation schedule, took good data, and
published science results
Cranked out 7 Stanford Ph.D.’s
Established an experimental astrophysics
presence at SLAC
Larry Wai, SLAC seminar
38
What’s left for USA?
Another Ph.D.
• Han Wen (Physics Ph.D.)
• Andrew Lee (Physics Ph.D.)
• John Hanson (Aero-Astro
Ph.D.)
• Alex Leubke (Aero-Astro
M.S/Engineering)
• Ganya Shabad (Physics Ph.D.)
• Kaice Reilly (App. Physics Ph.D.)
• Derek Tournear (Physics Ph.D.)
• Pablo Saz-Parkinson (Physics
Ph.D.)
• Daniel Engovatov (Physics Ph.D.,
in progress)
More papers:
• High frequency
power spectrum of
Cygnus X-1
• High frequency
QPO searches
(Circinus X-1, XTE
J1859+226)
• AGN studies
Larry Wai, SLAC seminar
39